Deep Learning-Based Classification Model for Suboptimal Conditions in Visible Light and Infrared Images of Photovoltaic Panels

E. Jiménez-Delgado; A. Méndez-Porras; J. Alfaro-Velasco; S. Campaña-Bastidas

doi:10.1007/978-3-031-93103-1_20

Deep Learning-Based Classification Model for Suboptimal Conditions in Visible Light and Infrared Images of Photovoltaic Panels

Universidad Nacional Abierta y a Distancia

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

The reduced energy output of photovoltaic systems can often be traced back to a decrease in the efficiency of photovoltaic modules due to various abnormal operating conditions, such as faults or malfunctions. Common issues affecting performance include: accumulation of dirt, hotspots, electrical faults, cracks, breakages, burns, and surface damage. The severity of the impact depends on numerous parameters and conditions. As the global demand for solar energy grows, so does the importance of automated defect detection in solar panels. Deep convolutional neural networks (CNNs) have demonstrated significant potential in addressing image classification tasks across a range of domains. In this study, we implement a CNN model to assess the surfaces of photovoltaic panels and identify defects. Initial results indicate that the model achieves an accuracy of 65% when applied to infrared thermal imagery and up to 85% for images captured under visible light conditions.

Original language	English
Title of host publication	Information Technology and Systems - ICITS 2025
Editors	Alvaro Rocha, Carlos Ferrás, Hiram Calvo
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	200-209
Number of pages	10
ISBN (Print)	9783031931024
DOIs	https://doi.org/10.1007/978-3-031-93103-1_20
State	Published - 2025
Event	International Conference on Information Technology and Systems, ICITS 2025 - Mexico City, Mexico Duration: 22 Jan 2025 → 25 Jan 2025

Publication series

Name	Lecture Notes in Networks and Systems
Volume	1449 LNNS
ISSN (Print)	2367-3370
ISSN (Electronic)	2367-3389

Conference

Conference	International Conference on Information Technology and Systems, ICITS 2025
Country/Territory	Mexico
City	Mexico City
Period	22/01/25 → 25/01/25

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 7 Affordable and Clean Energy

Keywords

CNN
Deep Learning
GANs
Neural Networks
Photovoltaic Panels
Suboptimal Conditions

Access to Document

10.1007/978-3-031-93103-1_20

Cite this

Jiménez-Delgado, E., Méndez-Porras, A., Alfaro-Velasco, J., & Campaña-Bastidas, S. (2025). Deep Learning-Based Classification Model for Suboptimal Conditions in Visible Light and Infrared Images of Photovoltaic Panels. In A. Rocha, C. Ferrás, & H. Calvo (Eds.), Information Technology and Systems - ICITS 2025 (pp. 200-209). (Lecture Notes in Networks and Systems; Vol. 1449 LNNS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-93103-1_20

Jiménez-Delgado, E. ; Méndez-Porras, A. ; Alfaro-Velasco, J. et al. / Deep Learning-Based Classification Model for Suboptimal Conditions in Visible Light and Infrared Images of Photovoltaic Panels. Information Technology and Systems - ICITS 2025. editor / Alvaro Rocha ; Carlos Ferrás ; Hiram Calvo. Springer Science and Business Media Deutschland GmbH, 2025. pp. 200-209 (Lecture Notes in Networks and Systems).

@inproceedings{b6588785879746ba9d50a4ed38b0b955,

title = "Deep Learning-Based Classification Model for Suboptimal Conditions in Visible Light and Infrared Images of Photovoltaic Panels",

abstract = "The reduced energy output of photovoltaic systems can often be traced back to a decrease in the efficiency of photovoltaic modules due to various abnormal operating conditions, such as faults or malfunctions. Common issues affecting performance include: accumulation of dirt, hotspots, electrical faults, cracks, breakages, burns, and surface damage. The severity of the impact depends on numerous parameters and conditions. As the global demand for solar energy grows, so does the importance of automated defect detection in solar panels. Deep convolutional neural networks (CNNs) have demonstrated significant potential in addressing image classification tasks across a range of domains. In this study, we implement a CNN model to assess the surfaces of photovoltaic panels and identify defects. Initial results indicate that the model achieves an accuracy of 65\% when applied to infrared thermal imagery and up to 85\% for images captured under visible light conditions.",

keywords = "CNN, Deep Learning, GANs, Neural Networks, Photovoltaic Panels, Suboptimal Conditions",

author = "E. Jim{\'e}nez-Delgado and A. M{\'e}ndez-Porras and J. Alfaro-Velasco and S. Campa{\~n}a-Bastidas",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.; International Conference on Information Technology and Systems, ICITS 2025 ; Conference date: 22-01-2025 Through 25-01-2025",

year = "2025",

doi = "10.1007/978-3-031-93103-1\_20",

language = "Ingl{\'e}s",

isbn = "9783031931024",

series = "Lecture Notes in Networks and Systems",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "200--209",

editor = "Alvaro Rocha and Carlos Ferr{\'a}s and Hiram Calvo",

booktitle = "Information Technology and Systems - ICITS 2025",

}

Jiménez-Delgado, E , Méndez-Porras, A , Alfaro-Velasco, J & Campaña-Bastidas, S 2025, Deep Learning-Based Classification Model for Suboptimal Conditions in Visible Light and Infrared Images of Photovoltaic Panels. in A Rocha, C Ferrás & H Calvo (eds), Information Technology and Systems - ICITS 2025. Lecture Notes in Networks and Systems, vol. 1449 LNNS, Springer Science and Business Media Deutschland GmbH, pp. 200-209, International Conference on Information Technology and Systems, ICITS 2025, Mexico City, Mexico, 22/01/25. https://doi.org/10.1007/978-3-031-93103-1_20

Deep Learning-Based Classification Model for Suboptimal Conditions in Visible Light and Infrared Images of Photovoltaic Panels. / Jiménez-Delgado, E.; Méndez-Porras, A.; Alfaro-Velasco, J. et al.
Information Technology and Systems - ICITS 2025. ed. / Alvaro Rocha; Carlos Ferrás; Hiram Calvo. Springer Science and Business Media Deutschland GmbH, 2025. p. 200-209 (Lecture Notes in Networks and Systems; Vol. 1449 LNNS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Deep Learning-Based Classification Model for Suboptimal Conditions in Visible Light and Infrared Images of Photovoltaic Panels

AU - Jiménez-Delgado, E.

AU - Méndez-Porras, A.

AU - Alfaro-Velasco, J.

AU - Campaña-Bastidas, S.

N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

PY - 2025

Y1 - 2025

N2 - The reduced energy output of photovoltaic systems can often be traced back to a decrease in the efficiency of photovoltaic modules due to various abnormal operating conditions, such as faults or malfunctions. Common issues affecting performance include: accumulation of dirt, hotspots, electrical faults, cracks, breakages, burns, and surface damage. The severity of the impact depends on numerous parameters and conditions. As the global demand for solar energy grows, so does the importance of automated defect detection in solar panels. Deep convolutional neural networks (CNNs) have demonstrated significant potential in addressing image classification tasks across a range of domains. In this study, we implement a CNN model to assess the surfaces of photovoltaic panels and identify defects. Initial results indicate that the model achieves an accuracy of 65% when applied to infrared thermal imagery and up to 85% for images captured under visible light conditions.

AB - The reduced energy output of photovoltaic systems can often be traced back to a decrease in the efficiency of photovoltaic modules due to various abnormal operating conditions, such as faults or malfunctions. Common issues affecting performance include: accumulation of dirt, hotspots, electrical faults, cracks, breakages, burns, and surface damage. The severity of the impact depends on numerous parameters and conditions. As the global demand for solar energy grows, so does the importance of automated defect detection in solar panels. Deep convolutional neural networks (CNNs) have demonstrated significant potential in addressing image classification tasks across a range of domains. In this study, we implement a CNN model to assess the surfaces of photovoltaic panels and identify defects. Initial results indicate that the model achieves an accuracy of 65% when applied to infrared thermal imagery and up to 85% for images captured under visible light conditions.

KW - CNN

KW - Deep Learning

KW - GANs

KW - Neural Networks

KW - Photovoltaic Panels

KW - Suboptimal Conditions

UR - https://www.scopus.com/pages/publications/105012921996

U2 - 10.1007/978-3-031-93103-1_20

DO - 10.1007/978-3-031-93103-1_20

M3 - Contribución a la conferencia

AN - SCOPUS:105012921996

SN - 9783031931024

T3 - Lecture Notes in Networks and Systems

SP - 200

EP - 209

BT - Information Technology and Systems - ICITS 2025

A2 - Rocha, Alvaro

A2 - Ferrás, Carlos

A2 - Calvo, Hiram

PB - Springer Science and Business Media Deutschland GmbH

T2 - International Conference on Information Technology and Systems, ICITS 2025

Y2 - 22 January 2025 through 25 January 2025

ER -

Jiménez-Delgado E , Méndez-Porras A , Alfaro-Velasco J, Campaña-Bastidas S. Deep Learning-Based Classification Model for Suboptimal Conditions in Visible Light and Infrared Images of Photovoltaic Panels. In Rocha A, Ferrás C, Calvo H, editors, Information Technology and Systems - ICITS 2025. Springer Science and Business Media Deutschland GmbH. 2025. p. 200-209. (Lecture Notes in Networks and Systems). doi: 10.1007/978-3-031-93103-1_20